Here we have the fact which the aspect of the glacier suggested. The first stake moves five inches a day more than the second, and the second nearly three inches a day more than the third. As surmised, therefore, the glacier is in a state of longitudinal compression, whereby a portion of it 1000 yards in length is shortened at the rate of eight inches a day. [Sidenote: STRUCTURE IN WHITE ICE-SEAMS.] In accordance with this result, the transverse undulations of the Glacier du Geant, described in the chapter upon Dirt-Bands, _shorten_ as they descend. A series of three of them measured along the axis of the glacier on the 6th of August, 1857, gave the following respective lengths:--955 links, 855 links, 770 links, the shortest undulation being the farthest from the origin of the undulations. This glacier then constitutes a vast ice-press, and enables us to test the explanation which refers the veined structure of the ice to pressure. The glacier itself is transversely laminated, as already stated; and in many cases a structure of extreme definition and beauty is developed in the compressed snow, which constitutes the seams of white ice. In 1857 I discovered a well-developed lenticular structure in some of these seams. In 1858 I again examined them. Clearing away the superficial portions with my axe, I found, drawn through the body of the seams, long lines of blue ice of exquisite definition; in fact, I had never seen the structure so delicately exhibited. The seams, moreover, were developed in portions of the white ice which were near the _centre_ of the glacier, and where consequently filamentous sliding was entirely out of the question. [Sidenote: PARTIAL SUMMARY.] PARTIAL SUMMARY. 1. Glaciers are derived from mountain snow, which has been consolidated to ice by pressure. 2. That pressure is competent to convert snow into ice has been proved by experiment. 3. The power of yielding to pressure diminishes as the mass becomes more compact; but it does not cease even when the substance has attained the compactness which would entitle it to be called ice. 4. When a sufficient depth of snow collects upon the earth's surface, the lower portions are squeezed out by the pressure of the superincumbent mass. If it rests upon a slope it will yield principally in the direction of the slope, and move downwards. 5. In addition to this, the whole mass slides bodily along its inclined bed, and leaves the traces of its sliding on the rocks over which it passes, grinding off their asperities, and marking them with grooves and scratches in the direction of the motion. 6. In this way the deposit of consolidated and unconsolidated snow which covers the higher portions of lofty mountains moves slowly down into an adjacent valley, through which it descends as a true glacier, partly by sliding and partly by the yielding of the mass itself.